Cross-border e-commerce data statistical management method for data processing and analysis

Abstract

The invention relates to the technical field of data processing, in particular to a cross-border e-commerce data statistical management method for data processing analysis, which comprises the following steps of: firstly, judging different e-commerce data except for the same order number, and independently storing the e-commerce data into an abnormal database; different order numbers and other corresponding data are stored in an order data storage database, then Pearson's correlation coefficients of different categories of e-commerce data in the order data storage database are calculated, and relevance between different categories of data is determined; and according to the calculated Pearson's correlation coefficient, abnormal e-commerce data in the order data storage database is judged to be deleted, abnormal data is independently stored and processed, and abnormal data in the order data storage database is deleted, so that the data volume is reduced, the data processing complexity is reduced, and the data processing efficiency is improved. Through calculation of the Pearson's correlation coefficient and abnormal data judgment based on the Pearson's correlation coefficient, problem data can be quickly positioned, and the efficiency of data analysis is improved.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a cross-border e-commerce data statistics management method for data processing and analysis. Background Art

[0002] In the statistical management process of cross-border e-commerce data, deduplication is a key pre-processing step. At present, the common deduplication technology is mainly based on order number comparison. Once the same order number is found, one of them will be deleted. However, this method has obvious limitations;

[0003] In actual operation, staff may enter order numbers incorrectly for various reasons. At this time, other information in the e-commerce data, such as product details, customer information, transaction time, etc., may be completely different. If one of the data is simply deleted based on the same order number, it is very likely to cause the loss of important e-commerce data, which will have a serious negative impact on subsequent data analysis, business decision-making, and customer service.

[0004] In addition, in the field of cross-border e-commerce, order numbers are usually arranged in order of size. When the order numbers are the same but other data are different, if the data with the same order numbers cannot be renumbered, these orders cannot be accurately distinguished and managed, which will not only cause confusion in data management, but also affect the smooth progress of business processes. In view of this, we propose a cross-border e-commerce data statistical management method for data processing and analysis. Summary of the invention

[0005] The purpose of the present invention is to solve the problem of directly deleting e-commerce data when the order numbers in the e-commerce data are the same but other data are different.

[0006] To achieve the above object, the present invention provides a cross-border e-commerce data statistics management method for data processing and analysis, comprising the following steps:

[0007] S1. Establish an order data storage database, and establish a connection with the cross-border e-commerce platform through the HTTP protocol to obtain and store e-commerce data;

[0008] S2. De-duplicate the e-commerce data in the order data storage database, and store the e-commerce data with the same order number but different other data in the abnormal database;

[0009] S3, using the Pearson correlation coefficient to calculate the correlation coefficients of different categories of all e-commerce data in the order data storage database after deduplication, and setting a correlation coefficient threshold, based on the correlation coefficient threshold and the correlation coefficient, determining whether there is a correlation between different categories in the e-commerce data, and if there is a correlation, feeding back the correlation in the e-commerce data to the staff, and determining abnormal data in the e-commerce data;

[0010] S4. Sort the e-commerce data in the order data storage database in the order of order numbers, and analyze whether there are discontinuous orders. If there are discontinuous order numbers, supplement the order numbers in descending order, and define the supplemented order numbers as supplementary numbers;

[0011] S5. Compare the supplementary number with the order number in the abnormal database to see whether the order numbers are the same, the number of identical order numbers, and the number of adjacent supplementary numbers are the same. If they are the same, replace the order number in the abnormal database with the supplementary number, and analyze again whether there is abnormal data therein.

[0012] As a further improvement of this technical solution, the working principle of establishing a connection between the order data storage database and the cross-border e-commerce platform is as follows:

[0013] The order data storage database sends a connection request to the cross-border e-commerce platform server address. After receiving the order data storage database connection request, the cross-border e-commerce platform server extracts the API key in the request, and the server searches and compares it in the API key list stored in itself;

[0014] If the server finds that the API key in the request is exactly the same as the API key stored in its own storage, it indicates that the authentication of the order data storage database is successful, and the e-commerce data in the cross-border e-commerce platform is obtained. Otherwise, the request is resent and verified.

[0015] As a further improvement of the technical solution, the e-commerce data in the order data storage database is deduplicated in S2, and the specific steps are as follows:

[0016] Step 1: sensing a query statement for querying an order number of e-commerce data stored in the order data storage database, and inputting the query statement into the order data storage database, traversing the e-commerce data stored in the order data storage database according to the query statement, calling out the order number corresponding to the query statement, that is, the order number of the e-commerce data in the order data storage database, and defining it as an order number set;

[0017] Step 2: randomly select an order number from the order number set based on the random selection logic, and compare it with the unselected order numbers based on the duplication determination logic to determine whether there is an identical number;

[0018] If there is no order number identical to the selected order number in the order number set, it is determined that the e-commerce data stored in the order data storage database is not repeated;

[0019] Step 3: If the same order number exists, call out other data of the same order number in the order data storage database; if the other data is still the same, delete the selected order number and the corresponding other data in the order data storage database, and continue to store the e-commerce data corresponding to the undeleted order number in the order data storage database;

[0020] If other data are different, the same order number and the e-commerce data corresponding to the order number are stored separately in the abnormal database;

[0021] An order number that has not been selected in the order number set is randomly selected again, and is compared with the order number that has not been compared in the order number set, until all order numbers in the order number set are compared.

[0022] As a further improvement of the technical solution, the expression corresponding to the random selection logic is as follows: Order number set , The Chinese Communist Party An order number is generated by a random number generator. , the range is The selected order number is , expressed as ,in Indicates that it is generated in arrive A function that generates a random number between .

[0023] As a further improvement of the technical solution, the expression corresponding to the duplication determination logic is as follows: For the selected order number and Another order number in , to determine whether it is repeated as , for all order numbers in the order data set, except All except The above judgment is expressed as and ,examine Is it true?

[0024] As a further improvement of this technical solution, to determine whether other data are the same, the logical expression based on data consistency is as follows: For other data corresponding to two identical order numbers and , other data and The order numbers are recorded as , other data consistency judgments are expressed as:

[0025] for ,judge Whether it is established can be expressed as a logical expression: ;

[0026] in Represents a logical AND operation. When the result of the expression is true, the other data are consistent; when the result is false, the other data are inconsistent.

[0027] As a further improvement of the technical solution, the working principle of calculating the correlation coefficient of the Pearson correlation coefficient in S3 is as follows:

[0028] Sense the e-commerce data after deduplication in the order data storage database, analyze the categories that are the same in all e-commerce data, and calculate the Pearson correlation coefficient between different categories;

[0029] The calculation formula of Pearson correlation coefficient is: ;

[0030] in and The first numerical values, and To select the average value of multiple e-commerce data, The number of e-commerce data selected.

[0031] As a further improvement of the technical solution, the working principle of determining whether there is a correlation between multiple e-commerce data in S3 is as follows:

[0032] If the correlation coefficient between the e-commerce data is greater than the maximum value of the correlation coefficient threshold, it is determined that there is a correlation between the e-commerce order data, which is a positive linear correlation. If the correlation coefficient is less than the minimum value of the correlation coefficient threshold, it is determined that there is a correlation between the e-commerce order data, which is a negative linear correlation. Otherwise, it is determined that there is no linear correlation between the e-commerce order data.

[0033] As a further improvement of the present technical solution, the working principle of determining whether there is a correlation between different categories in the e-commerce data in S3 is as follows: after the correlation analysis between the e-commerce data is completed, randomly select more than 2 e-commerce data through random selection logic again, calculate the correlation coefficient between the different categories of data in the selected e-commerce data, and again determine whether there is a correlation between the different categories of data through the correlation coefficient threshold. If there is a correlation in the analysis process of all e-commerce data, when there is no correlation at this time, the randomly selected e-commerce data is defined as an abnormal data set;

[0034] E-commerce data is randomly selected again. During the selection process, any e-commerce data in the abnormal data set is added. If the correlation coefficient threshold is used to determine that there is a correlation between data of different categories, the added e-commerce data is removed from the abnormal data set. Otherwise, the added e-commerce data is determined to be abnormal data and deleted from the order data storage database.

[0035] As a further improvement of the technical solution, the order numbers are sorted in the order of size in S4. The basic idea is to select a base order number, put the order numbers smaller than the base order number on the left, and the order numbers larger than the base order number on the right, and then sort the left and right parts respectively;

[0036] In S4, the sorted order number sequence is traversed and checked, and the difference between two adjacent order numbers is calculated. If the difference between the adjacent order numbers is greater than 1, it is determined that the two adjacent numbers are not continuous.

[0037] Compared with the prior art, the present invention has the following beneficial effects:

[0038] In the cross-border e-commerce data statistical management method for data processing and analysis, different e-commerce data except for the same order number are first determined, and such e-commerce data are stored separately in an abnormal database, different order numbers and corresponding other data are stored in an order data storage database, and then the Pearson correlation coefficients of different categories of e-commerce data in the order data storage database are calculated, and the correlation between different categories of data is determined, and the abnormal e-commerce data in the order data storage database is determined and deleted according to the calculated Pearson correlation coefficient;

[0039] The separate storage and processing of abnormal data, as well as the deletion of abnormal data in the order data storage database, reduces the amount of data and the complexity of data processing. The calculation of the Pearson correlation coefficient and the abnormal data determination based on it can quickly locate the problem data and improve the efficiency of data analysis;

[0040] Finally, by sorting the e-commerce data according to its order number, supplementing the missing order numbers during the sorting process, and renumbering the e-commerce data in the abnormal database, the Pearson correlation coefficient is used to calculate whether the renumbered e-commerce data is abnormal data. This encourages companies to no longer focus solely on order numbers when deduplicating, but to comprehensively consider other e-commerce data and explore more comprehensive and accurate deduplication algorithms, such as combining multiple data features for comparison, to improve the accuracy of deduplication and reduce the risk of data loss. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 It is a flow chart of the overall steps of the present invention. DETAILED DESCRIPTION

[0042] The following will be combined with the accompanying drawings in the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0043] A cross-border e-commerce data statistical management method for data processing and analysis includes the following steps:

[0044] S1. Establish an order data storage database, and establish a connection with the cross-border e-commerce platform through the HTTP protocol to obtain and store e-commerce data. The specific working principle is as follows:

[0045] E-commerce data includes: order data (including order number, order time, customer information, product details, payment amount, etc.), product data (product ID, name, description, price, inventory, etc.), customer evaluation data, etc.;

[0046] Before the order data storage database establishes a connection with the cross-border e-commerce platform, the cross-border e-commerce platform provides a server address and API key. The server address is the target network location for the order data storage database to communicate with the platform, and the API key is key information for identity authentication. It is unique, just like a key. Only the order data storage database holding the correct key can legally access the platform data.

[0047] The order data storage database sends a connection request to the cross-border e-commerce platform server address (the connection request includes the API key and the e-commerce data to be obtained). After receiving the order data storage database connection request, the cross-border e-commerce platform server extracts the API key in the request, and searches and compares it in its own stored API key list (the API key list is a database or similar data storage structure that stores all legal API keys). If the server finds that the API key in the request is exactly the same as an API key stored in its own storage, it indicates that the identity authentication of the order data storage database is successful, and the e-commerce data in the cross-border e-commerce platform is obtained. Otherwise, the request is resent and verified.

[0048] S2. De-duplication of e-commerce data in the order data storage database. The specific steps are as follows:

[0049] Step 1: Sense the query statement for querying the order number of the e-commerce data stored in the order data storage database (the query statement is an instruction for retrieving, filtering, and operating data from the e-commerce data stored in the order data storage database), and input the query statement into the order data storage database, traverse the e-commerce data stored in the order data storage database according to the query statement, and call out the order number corresponding to the query statement, which is the order number of the e-commerce data in the order data storage database, and is defined as an order number set. Since the order number is set as a key field that uniquely identifies each order in the entire e-commerce data system, the order number can ensure that each order has a unique identity;

[0050] Step 2: randomly select an order number from the order number set based on the random selection logic, and compare it with the unselected order numbers based on the duplication determination logic to determine whether there is an identical number;

[0051] The expression corresponding to the random selection logic is as follows: Order number set , The Chinese Communist Party An order number is generated by a random number generator. , the range is The selected order number is , expressed as ,in Indicates that it is generated in arrive A function of random numbers between ;

[0052] The expression corresponding to the duplicate determination logic is as follows: For the selected order number and Another order number in , to determine whether it is repeated as , for all order numbers in the order data set, except All except The above judgment is expressed as and ,examine whether it is established;

[0053] If there is no order number identical to the selected order number in the order number set, it is determined that the e-commerce data stored in the order data storage database is not repeated.

[0054] In cross-border e-commerce, e-commerce data management is extremely critical. Business expansion has led to a surge in the amount of e-commerce data. Due to manual entry errors, system failures, data synchronization deviations, etc., duplicate order numbers often appear in the order data storage database;

[0055] If e-commerce data duplication is determined based on this alone, valuable data may be mistakenly deleted. For example, if different customers place orders at similar times and the same order number is generated due to a system failure, but other data such as the product and delivery address are different, simply duplicating the order number will result in loss of business information, affecting inventory, logistics and customer service, damaging customer experience, and even leading to customer loss.

[0056] On the other hand, it interferes with data analysis results. When analyzing sales, customer behavior, and market strategies, inaccurate e-commerce data can bias the results and mislead decision-making. For example, mistakenly deleting valid order data can lead to a low evaluation of regional sales performance, which can reduce corporate investment and miss market growth opportunities. At the same time, it increases the workload and cost of data processing, requiring more manpower and time to investigate and correct errors, and reduces operational efficiency.

[0057] Step 3: If the same order number exists, call out other data of the same order number in the order data storage database (other data refers to all e-commerce data except the order number); if other data is still the same, delete the selected order number and other corresponding data in the order data storage database, and continue to store the e-commerce data corresponding to the undeleted order number in the order data storage database;

[0058] The logical expression for judging whether other data are the same is as follows: For other data corresponding to two identical order numbers, and , other data and The order numbers are recorded as , other data consistency judgments are expressed as:

[0059] for ,judge Whether it is established can be expressed as a logical expression: ;

[0060] in Indicates a logical AND operation. When the result of the expression is true, the other data are consistent; when the result is false, the other data are inconsistent.

[0061] If other data are different, the same order number and the e-commerce data corresponding to the order number are stored separately in the abnormal database;

[0062] An order number that has not been selected in the order number set is randomly selected again, and is compared with the order number that has not been compared in the order number set, until all order numbers in the order number set are compared.

[0063] S3. Use the Pearson correlation coefficient to calculate the correlation coefficients of different categories of all e-commerce data in the order data storage database after deduplication, and set a correlation coefficient threshold (the correlation coefficient threshold refers to the general standard for judging the correlation of different indicators in the industry. In the e-commerce industry, it is generally believed that when the absolute value of the correlation coefficient between commodity price and sales volume is greater than 0.6, the two are strongly correlated. This 0.6 is an empirical value based on a large amount of practice and research in the industry). Based on the correlation coefficient threshold and the correlation coefficient, determine whether there is a correlation between multiple e-commerce data. If there is a correlation, feedback the correlation in the e-commerce data to the staff;

[0064] And according to the correlation between e-commerce data, delete the e-commerce data that does not meet the correlation in the order data storage database;

[0065] Pearson Correlation Coefficient Calculating the correlation coefficient works as follows:

[0066] Sense the e-commerce data after deduplication in the order data storage database, analyze the categories that are the same in all e-commerce data, and calculate the Pearson correlation coefficient between different categories;

[0067] The calculation formula of Pearson correlation coefficient is: ;

[0068] in and The first numerical values, and To select the average value of multiple e-commerce data, The number of e-commerce data selected;

[0069] The working principle of determining whether there is a correlation between multiple e-commerce data is as follows:

[0070] If the correlation coefficient between e-commerce data is greater than the maximum value of the correlation coefficient threshold, it is determined that there is a correlation between the e-commerce order data, which is a positive linear correlation. If the correlation coefficient is less than the minimum value of the correlation coefficient threshold, it is determined that there is a correlation between the e-commerce order data, which is a negative linear correlation. Otherwise, it is determined that there is no linear correlation between the e-commerce order data.

[0071] Positive linear correlation: This means that when one data in the e-commerce order increases, the other data will also increase, and the increasing trend shows a linear characteristic. When inventory and price show a positive linear correlation, it may mean that as inventory increases, prices will also increase accordingly, indicating that there may be some kind of inherent positive relationship between them, which may be due to factors such as inventory costs.

[0072] Negative linear correlation: When one data in an e-commerce order increases, the other data decreases at roughly the same rate. For example, in some promotional activities, prices may drop while sales (which can be reflected indirectly from changes in inventory) increase. At this time, changes in price and inventory may show a negative linear correlation.

[0073] By performing deduplication operations on e-commerce data, the uniqueness of the acquired e-commerce data is ensured. After this step, duplicate e-commerce data is removed, data redundancy is reduced, and subsequent analysis is more efficient. In addition, since the Pearson correlation coefficient between different categories is calculated through all e-commerce data, the abnormal data may not conform to normal business logic or data rules due to its own particularity. However, due to their small proportion, in the process of calculating the correlation coefficient, their contribution to the overall calculation result is relatively small, and will not significantly distort or change the correlation coefficient value calculated based on normal data, will not cause a large deviation in the correlation coefficient, and will not cause substantial interference to the analysis of the correlation between different categories;

[0074] After the correlation analysis between e-commerce data is completed, more than 2 e-commerce data are randomly selected again through random selection logic, and the correlation coefficients between different categories of data in the selected e-commerce data are calculated. The correlation coefficient threshold is used again to determine whether there is a correlation between different categories of data. If there is a correlation in the analysis process of all e-commerce data, and there is no correlation at this time, the randomly selected e-commerce data is defined as an abnormal data set;

[0075] Then randomly select e-commerce data again. During the selection process, add any e-commerce data in the abnormal data set. If the correlation coefficient threshold is used to determine that there is a correlation between data of different categories, the added e-commerce data is removed from the abnormal data set. Otherwise, the added e-commerce data is determined to be abnormal data and deleted from the order data storage database.

[0076] This process can gradually optimize the data quality in the database by continuously screening and adjusting the abnormal data set, and finally deleting the abnormal data from the order data storage database. After multiple screening and verification, it ensures that the data remaining in the database is more consistent with the real correlation between the data, avoids the misleading of subsequent data analysis and decision support by abnormal data, and provides a high-quality data foundation for more accurate business decisions.

[0077] S4. Sort the e-commerce data stored in the order data database in the order of order numbers;

[0078] Sort by order number. The basic idea is to select a base order number, put the order numbers smaller than the base order number on the left, and the order numbers larger than the base order number on the right, and then sort the left and right parts separately;

[0079] For separately stored order numbers and ,like , then exchange and The position of the order numbers is determined by multiple comparisons to determine the order of the separately stored order numbers;

[0080] And analyze whether there are discontinuous orders. If there are discontinuous order numbers, supplement the order numbers in descending order, and define the supplemented order numbers as supplementary numbers;

[0081] The sorted order number sequence is traversed and checked. The difference between two adjacent order numbers is calculated. If the difference between two adjacent order numbers is greater than 1, the two adjacent numbers are judged to be discontinuous. For the sequence [1003, 1001, 1000], after sorting, it becomes [1003, 1001, 1000]. The difference between the adjacent elements 1003 and 1001 is 2, which is judged to be discontinuous.

[0082] When discontinuous order numbers are found, a supplementary number is generated based on the difference between adjacent numbers. For 1003 and 1001 in the above example, the supplementary number is 1002. The generation of supplementary numbers is calculated based on the sequence and the difference, filling in the discontinuous gaps from large to small.

[0083] S5. Compare the supplementary number with the order number in the abnormal database to see whether the order numbers are the same, the number of identical order numbers, and the number of adjacent supplementary numbers are the same. If they are the same, replace the order number in the abnormal database with the supplementary number, and analyze again whether there is abnormal data therein.

[0084] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments. The above embodiments and descriptions are only preferred examples of the present invention and are not intended to limit the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. A cross-border e-commerce data statistical management method for data processing and analysis, characterized in that: The following steps are involved: S1. Establish an order data storage database, and establish a connection with the cross-border e-commerce platform through the HTTP protocol to obtain and store e-commerce data; S2. De-duplicate the e-commerce data in the order data storage database, and store the e-commerce data with the same order number but different other data in the abnormal database; S3, using the Pearson correlation coefficient to calculate the correlation coefficients of different categories of all e-commerce data in the order data storage database after deduplication, and setting a correlation coefficient threshold, based on the correlation coefficient threshold and the correlation coefficient, determining whether there is a correlation between different categories in the e-commerce data, and if there is a correlation, feeding back the correlation in the e-commerce data to the staff, and determining abnormal data in the e-commerce data; S4. Sort the e-commerce data in the order data storage database in the order of order numbers, and analyze whether there are discontinuous orders. If there are discontinuous order numbers, supplement the order numbers in descending order, and define the supplemented order numbers as supplementary numbers; S5. Compare the supplementary number with the order number in the abnormal database to see whether the order numbers are the same, the number of identical order numbers, and the number of adjacent supplementary numbers are the same. If they are the same, replace the order number in the abnormal database with the supplementary number, and analyze again whether there is abnormal data therein.

2. The cross-border e-commerce data statistics management method for data processing and analysis according to claim 1 is characterized by: The working principle of establishing a connection between the order data storage database and the cross-border e-commerce platform is as follows: The order data storage database sends a connection request to the cross-border e-commerce platform server address. After receiving the order data storage database connection request, the cross-border e-commerce platform server extracts the API key in the request, and the server searches and compares it in the API key list stored in itself; If the server finds that the API key in the request is exactly the same as the API key stored in its own storage, it indicates that the authentication of the order data storage database is successful, and the e-commerce data in the cross-border e-commerce platform is obtained. Otherwise, the request is resent and verified.

3. The cross-border e-commerce data statistics management method for data processing and analysis according to claim 2 is characterized by: In S2, the e-commerce data in the order data storage database is deduplicated. The specific steps are as follows: Step 1: sensing a query statement for querying an order number of e-commerce data stored in the order data storage database, and inputting the query statement into the order data storage database, traversing the e-commerce data stored in the order data storage database according to the query statement, calling out the order number corresponding to the query statement, that is, the order number of the e-commerce data in the order data storage database, and defining it as an order number set; Step 2: randomly select an order number from the order number set based on the random selection logic, and compare it with the unselected order numbers based on the duplication determination logic to determine whether there is an identical number; If there is no order number identical to the selected order number in the order number set, it is determined that the e-commerce data stored in the order data storage database is not repeated; Step 3: If the same order number exists, call out other data of the same order number in the order data storage database; if the other data is still the same, delete the selected order number and the corresponding other data in the order data storage database, and continue to store the e-commerce data corresponding to the undeleted order number in the order data storage database; If other data are different, the same order number and the e-commerce data corresponding to the order number are stored separately in the abnormal database; An order number that has not been selected in the order number set is randomly selected again, and is compared with the order number that has not been compared in the order number set, until all order numbers in the order number set are compared.

4. The cross-border e-commerce data statistics management method for data processing and analysis according to claim 3 is characterized by: The expression corresponding to the random selection logic is as follows: Order number set , The Chinese Communist Party An order number is generated by a random number generator. , the range is The selected order number is , expressed as ,in Indicates that it is generated in arrive A function that generates a random number between .

5. The cross-border e-commerce data statistics management method for data processing and analysis according to claim 4 is characterized by: The expression corresponding to the duplication determination logic is as follows: For the selected order number and Another order number in , to determine whether it is repeated as , for all order numbers in the order data set, except All except The above judgment is expressed as and ,examine Is it true? 6. The cross-border e-commerce data statistics management method for data processing and analysis according to claim 5 is characterized by: The logical expression for judging whether other data are the same is as follows: For other data corresponding to two identical order numbers, and , other data and The order numbers are recorded as , other data consistency judgments are expressed as: for ,judge Whether it is established can be expressed as a logical expression: ; in Represents a logical AND operation. When the result of the expression is true, the other data are consistent; when the result is false, the other data are inconsistent.

7. The cross-border e-commerce data statistics management method for data processing and analysis according to claim 3 is characterized by: The working principle of calculating the correlation coefficient of the Pearson correlation coefficient in S3 is as follows: Sense the e-commerce data after deduplication in the order data storage database, analyze the categories that are the same in all e-commerce data, and calculate the Pearson correlation coefficient between different categories; The calculation formula of Pearson correlation coefficient is: ; in and The first numerical values, and To select the average value of multiple e-commerce data, The number of e-commerce data selected.

8. The cross-border e-commerce data statistics management method for data processing and analysis according to claim 7 is characterized by: The working principle of determining whether there is a correlation between multiple e-commerce data in S3 is as follows: If the correlation coefficient between the e-commerce data is greater than the maximum value of the correlation coefficient threshold, it is determined that there is a correlation between the e-commerce order data, which is a positive linear correlation. If the correlation coefficient is less than the minimum value of the correlation coefficient threshold, it is determined that there is a correlation between the e-commerce order data, which is a negative linear correlation. Otherwise, it is determined that there is no linear correlation between the e-commerce order data.

9. The cross-border e-commerce data statistics management method for data processing and analysis according to claim 8 is characterized by: The working principle of determining whether there is a correlation between different categories in the e-commerce data in S3 is as follows: after the correlation analysis between the e-commerce data is completed, randomly select more than 2 e-commerce data through random selection logic again, calculate the correlation coefficient between different categories of data in the selected e-commerce data, and again determine whether there is a correlation between different categories of data through the correlation coefficient threshold. If there is a correlation in the analysis process of all e-commerce data, when there is no correlation at this time, the randomly selected e-commerce data is defined as an abnormal data set; E-commerce data is randomly selected again. During the selection process, any e-commerce data in the abnormal data set is added. If the correlation coefficient threshold is used to determine that there is a correlation between data of different categories, the added e-commerce data is removed from the abnormal data set. Otherwise, the added e-commerce data is determined to be abnormal data and deleted from the order data storage database.

10. The cross-border e-commerce data statistics management method for data processing and analysis according to claim 1 is characterized by: In S4, the order numbers are sorted in order of size. The basic idea is to select a base order number, put the order numbers smaller than the base order number on the left, and the order numbers larger than the base order number on the right, and then sort the left and right parts respectively; In S4, the sorted order number sequence is traversed and checked, and the difference between two adjacent order numbers is calculated. If the difference between the adjacent order numbers is greater than 1, it is determined that the two adjacent numbers are not continuous.

Images